Method, system and equipment for identifying calling abnormity of production data of pumped storage power station and storage medium

Abstract

The invention discloses a method, a system and equipment for identifying calling abnormity of production data of a pumped storage power station and a storage medium. The method comprises the following steps: acquiring a data calling record of pumped storage power station production data called by an application through an interface mode, and generating a feature set corresponding to the production data based on the data calling record; calculating the correlation degree between the production data according to the similarity of the feature sets of the production data, and dividing the production data of which the correlation degree is greater than a threshold value into similar production data sets; constructing and training to obtain an interface feedback model, selecting a target calling record, and calculating an expected interface feedback value of the target calling record after a period of time; and calculating an actual interface feedback value of the target calling record after the period of time, and judging whether to carry out abnormity reminding on a calling behavior corresponding to the target calling record or not by comparing the difference degree between the actual interface feedback value of the target calling record and an expected interface feedback value. The method is high in accuracy.

Description

Technical Field

[0001] The present invention relates to anomaly recognition technology, and in particular to a method, device and storage medium for identifying abnormalities in calling production data of a pumped storage power station. Background Art

[0002] With the rapid development of the pumped storage industry, the generation of a large amount of production data is of great value in improving energy efficiency, optimizing production management, supporting scientific decision-making and promoting industrial upgrading. When the production data of the pumped storage power station production system is called by downstream applications, there may be abnormal calling behaviors, which may cause system abnormalities. Therefore, identifying abnormal calling behaviors is important for the safe operation of pumped storage power stations. Summary of the invention

[0003] In view of the problems existing in the prior art, the purpose of the present invention is to provide a method, device and storage medium for identifying abnormal call of production data of a pumped storage power station.

[0004] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0005] A method for identifying abnormal call of production data of a pumped storage power station, comprising:

[0006] Step S100: obtaining data call records of the application calling the production data of the pumped-storage power station through the interface, and obtaining high-frequency call fields in the corresponding production data based on the data call records, and using the high-frequency call fields together with the production data itself as a feature set of the corresponding production data; wherein each data call record is used to record all called data areas and call times in the production data of a production system of the pumped-storage power station;

[0007] Step S200: calculating the correlation between each production data according to the similarity of the feature set of the production data, and classifying the production data with a correlation greater than a threshold into a similar production data set;

[0008] Step S300: Calculate the calling frequency and calling data volume of production data at each moment according to the called data area and calling time recorded in the data calling record;

[0009] Step S400: constructing and training an interface feedback model, wherein the interface feedback model calculates the interface feedback value of the corresponding data call record according to the call frequency, call data volume and correlation between production data at each moment, wherein the interface feedback value is positively correlated with the call frequency, call data volume and correlation between production data;

[0010] Step S500: Set the data call record whose generation time is closest to the current moment as the target call record, and calculate the real-time interface feedback value of the target call record according to the interface feedback model, obtain the data call record whose production data corresponding to the target call record belongs to the same production data set, and calculate the probability that the target call record will subsequently generate the same type of data call record according to these data call records, as well as the expected interface feedback value of the target call record after a period of time;

[0011] Step S600: Calculate the actual interface feedback value of the target call record after the period of time according to the interface feedback model, and determine whether to issue an abnormal reminder for the call behavior corresponding to the target call record by comparing the difference between the actual interface feedback value of the target call record and the expected interface feedback value.

[0012] Furthermore, the step S100 specifically includes:

[0013] Step S101: Acquire all called data with call records in the production data stored in the production data storage service system, count the occurrence frequency of each field of all called data, and store each field and the corresponding occurrence frequency in the data content database; wherein the production data and data call records of the pumped storage power station are stored in the production data storage service system;

[0014] Step S102: extract all data call records from the production data storage service system, divide the corresponding production data of each data call record into several information blocks, compare the fields in each information block with the fields stored in the data content database, if the fields of the information block and the fields stored in the data content database have the same fields, select the field with the highest frequency of occurrence from the same fields as the characteristic field of the corresponding information block, and use the information block itself as the characteristic information of the corresponding information block; if there are no identical fields, directly use the information block itself as the characteristic information of the corresponding information block;

[0015] Step S103: The set of characteristic fields and characteristic information of all information blocks of the production data corresponding to the data call record is taken as its characteristic set.

[0016] Furthermore, the step S200 specifically includes:

[0017] Step S201: For any data call record i and any data call record j, respectively calculate the number of identical feature fields H in the feature field set of production data corresponding to the data call record i and the feature field set of production data corresponding to the data call record j ij , and the number of identical feature fields H ijDivide by the maximum number of fields in the two feature field sets to get the proportion of the same feature fields in the production data corresponding to data call record i and data call record j X ij ;

[0018] Step S202: for any data call record i and any data call record j, calculate the similarity between each feature information of the production data corresponding to the data call record i and each feature information of the production data corresponding to the data call record j, and if the similarity is less than the similarity threshold β, assign the corresponding similarity to 0;

[0019] Step S203: For any data call record i and any data call record j, the similarity S of the feature information set of the production data corresponding to the data call record i and the data call record j is calculated according to the following formula: ij ;

[0020]

[0021] In the formula, (S ij ) ab N represents the similarity between the ath feature information of the production data corresponding to the data call record i and the bth feature information of the production data corresponding to the data call record j, i 、N j They represent the number of characteristic information of production data corresponding to data call records i and j, respectively, and Max() is the maximum value function;

[0022] Step S204: If X ij <γ, then the correlation G between the data call records i and j corresponding to the production data ij Assign a value of 0; if X ij ≥γ, then G ij =X ij ×(1+S ij ), where γ is the preset proportion threshold;

[0023] Step S205: If G ij ≠0, the production data corresponding to the data call records i and j are classified into a category of data, and the data call record i and the data call record j are generated into a set of data call records of the same category.

[0024] Furthermore, the step S400 specifically includes:

[0025] Step S401: constructing an interface feedback model:

[0026]

[0027] In the formula, (F i ) trepresents the interface feedback value of the current data call record i at time t, y is the coefficient to be trained, (P i ) t , (Q i ) t They represent the calling frequency and calling data volume of the production data corresponding to the data calling record i at time t, G ki Indicates the correlation between the production data corresponding to data call record k and data call record i, where the index number of the data call record is assigned in sequence starting from 1 according to the generation time, the generation time of data call record k is before the generation time of data call record i, and t represents the tth unit time after the production time;

[0028] Step S402: Select several data call records from the production data storage service system as samples, and for each sample, obtain other data call records whose generation time is before the sample, and obtain the correlation between the production data corresponding to these data call records and the production data corresponding to the current data call record;

[0029] Step S403: Calculate the difference between the total amount of data at any two adjacent moments of each sample, where the total amount of data is the product of the call frequency and the call data amount;

[0030] Step S404: With the goal of making the difference in the total amount of data between any two adjacent moments of each sample consistent with the change in the corresponding interface feedback value, the interface feedback model is trained to obtain the coefficients to be trained y, and the training of the interface feedback model is completed.

[0031] Furthermore, the step S500 specifically includes:

[0032] Step S501: Take the data call record with the closest generation time to the current time as the target call record, and obtain the call frequency P of the production data corresponding to the target call record g And data call volume Q g , and calculate the real-time interface feedback value of the target call record according to the following formula based on the interface feedback model

[0033] Step S502: Obtain the feature set T of the production data corresponding to the target call record g =(W g ,M g ), where W g The characteristic field set of the target call record corresponding to the production data, M g The characteristic information set of the corresponding production data is recorded for the target call, and W g Compare it with the characteristic field set of production data corresponding to all other data call records, and compare it with W gThe same target call record is set as a reference record of the target call record;

[0034] Step S503: among all reference records of the target call record, the reference records whose production data belong to the same type of production data set are divided into one set to obtain a plurality of reference record sets;

[0035] Step S504: Obtain the number of data call records whose correlation degree with the production data corresponding to the target call record is not 0 in the production data storage service system. g , if 1+U g <U r , then the corresponding reference record set r is set as the valid set, and the ratio of the number of valid sets to the number of reference record sets is calculated as the probability η that the target call record will subsequently generate the same type of data call record; where U r is the number of reference records in the reference record set r, and the same type of data call record represents the data call record in the same type of data record set;

[0036] Step S505: M g Perform similarity comparison with the feature information set of all data call records in all valid sets, select the data call record with the greatest similarity as the most similar data call record, and extract the data call record whose corresponding production data's most recent call time is after the most similar data call record from the valid set where the most similar data call record is located, and use the interval between the most recent call times of the production data of these two data call records as the expected time interval Δt for the subsequent generation of similar data call records by the target call record;

[0037] Step S506: According to the expected time interval Δt and the probability η, the expected interface feedback value of the target call record after the expected time interval Δt has passed is calculated according to the following formula:

[0038]

[0039] In the formula, (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt.

[0040] Furthermore, the step S600 specifically includes:

[0041] Step S601: Obtain the calling frequency P of the production data corresponding to the target calling record after the expected time interval Δt now and the amount of data called Q now , and calculate the actual interface feedback value (F) of the target call record after the expected time interval Δt according to the interface feedback model g )ac ;

[0042] Step S602: determining whether a new data call record of the same type is generated after the target call record in the production data storage service system;

[0043] Step S603: If a new data call record of the same type is generated, and Then an abnormal reminder is given for the calling behavior recorded in the target calling record, wherein (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt. Indicates the real-time interface feedback value of the target call record.

[0044] A system for identifying abnormal calls of production data of a pumped storage power station, comprising:

[0045] A feature extraction module is used to obtain data call records of the application calling the production data of the pumped storage power station through the interface, and based on the data call records, obtain the high-frequency call fields in the corresponding production data, and use them together with the production data itself as the feature set of the corresponding production data; wherein each data call record is used to record all called data areas and call times in the production data of a production system of the pumped storage power station;

[0046] A classification module is used to calculate the correlation between each production data according to the similarity of the feature set of the production data, and classify the production data with a correlation greater than a threshold into a similar production data set;

[0047] The calling feature calculation module is used to calculate the calling frequency and calling data volume of the production data at each moment according to the called data area and calling time recorded in the data calling record;

[0048] A model training module is used to construct and train an interface feedback model, wherein the interface feedback model calculates the interface feedback value of the corresponding data call record according to the call frequency, call data volume and correlation between production data at each moment, wherein the interface feedback value is positively correlated with the call frequency, call data volume and correlation between production data;

[0049] The expected interface feedback calculation module is used to set the data call record with the closest generation time to the current time as the target call record, and calculate the real-time interface feedback value of the target call record according to the interface feedback model, obtain the data call record whose production data corresponding to the target call record belongs to the same production data set, and calculate the probability of the target call record generating the same type of data call record later according to these data call records, as well as the expected interface feedback value of the target call record after a period of time;

[0050] The abnormality identification module is used to calculate the actual interface feedback value of the target call record after the period of time according to the interface feedback model, and determine whether to issue an abnormal reminder for the call behavior corresponding to the target call record by comparing the difference between the actual interface feedback value of the target call record and the expected interface feedback value.

[0051] Furthermore, the feature extraction module specifically includes:

[0052] A high-frequency call field statistics unit is used to obtain all called data with call records in the production data stored in the production data storage service system, count the occurrence frequency of each field of all called data, and store each field and the corresponding occurrence frequency in the data content database; wherein the production data and data call records of the pumped storage power station are stored in the production data storage service system;

[0053] The feature extraction unit is used to extract all data call records from the production data storage service system. For each data call record, the corresponding production data is divided into several information blocks, and each field in each information block is compared with the fields stored in the data content database. If the fields of the information block and the fields stored in the data content database are identical, the field with the highest frequency of occurrence is selected from the identical fields as the feature field of the corresponding information block, and the information block itself is used as the feature information of the corresponding information block. If there are no identical fields, the information block itself is directly used as the feature information of the corresponding information block.

[0054] The feature forming unit is used to take the set of feature fields and feature information of all information blocks of the production data corresponding to the data call record as its feature set.

[0055] Furthermore, the classification module specifically includes:

[0056] The same field calculation unit is used to calculate the number of the same feature fields in the feature field set of the production data corresponding to the data call record i and the feature field set of the production data corresponding to the data call record j for any data call record i and any data call record j, respectively. ij , and the number of identical feature fields Hij Divide by the maximum number of fields in the two feature field sets to get the proportion of the same feature fields in the production data corresponding to data call record i and data call record j X ij ;

[0057] The information block similarity calculation unit is used to calculate the similarity between each feature information of the production data corresponding to the data call record i and each feature information of the production data corresponding to the data call record j for any data call record i and any data call record j, and if the similarity is less than a similarity threshold β, the corresponding similarity is assigned to 0;

[0058] The feature information similarity calculation unit is used to calculate the similarity S of the feature information set of the production data corresponding to the data call record i and the data call record j according to the following formula for any data call record i and any data call record j ij ;

[0059]

[0060] In the formula, (S ij ) ab N represents the similarity between the ath feature information of the production data corresponding to the data call record i and the bth feature information of the production data corresponding to the data call record j, i 、N j They represent the number of characteristic information of production data corresponding to data call records i and j, respectively, and Max() is the maximum value function;

[0061] The correlation calculation unit is used if X ij <γ, then the correlation G between the data call records i and j corresponding to the production data ij Assign a value of 0; if X ij ≥γ, then G ij =X ij ×(1+S ij ), where γ is the preset proportion threshold;

[0062] Classification unit, used if G ij ≠0, the production data corresponding to the data call records i and j are classified into a category of data, and the data call record i and the data call record j are generated into a set of data call records of the same category.

[0063] Furthermore, the model training module specifically includes:

[0064] The model building unit, used to build the interface feedback model is:

[0065]

[0066] In the formula, (Fi ) t represents the interface feedback value of the current data call record i at time t, y is the coefficient to be trained, (P i ) t , (Q i ) t They represent the calling frequency and calling data volume of the production data corresponding to the data calling record i at time t, G ki Indicates the correlation between the production data corresponding to data call record k and data call record i, where the index number of the data call record is assigned in sequence starting from 1 according to the generation time, the generation time of data call record k is before the generation time of data call record i, and t represents the tth unit time after the production time;

[0067] A sample selection unit is used to select a number of data call records from the production data storage service system as samples, and for each sample, obtain other data call records whose generation time is before the sample, and obtain the correlation between the production data corresponding to these data call records and the production data corresponding to the current data call record;

[0068] A data total amount calculation unit, used to calculate the difference between the total amount of data of each sample at any two adjacent moments, wherein the total amount of data is the product of the call frequency and the call data amount;

[0069] The model training unit is used to train the interface feedback model with the goal of making the change in the difference in the total amount of data between any two adjacent moments of each sample consistent with the change in the corresponding interface feedback value, obtain the coefficient y to be trained, and complete the training of the interface feedback model.

[0070] Furthermore, the expected interface feedback calculation module specifically includes:

[0071] The real-time interface sensing calculation unit is used to take the data call record with the closest generation time to the current time as the target call record, and obtain the call frequency P of the target call record corresponding to the production data g And data call volume Q g , and calculate the real-time interface feedback value of the target call record according to the following formula based on the interface feedback model

[0072] Reference record acquisition unit, used to obtain the feature set T of the production data corresponding to the target call record g =(W g ,M g ), where W g The characteristic field set of the target call record corresponding to the production data, M g The characteristic information set of the corresponding production data is recorded for the target call, and W gCompare it with the characteristic field set of production data corresponding to all other data call records, and compare it with W g The same target call record is set as a reference record of the target call record;

[0073] A reference record set acquisition unit is used to divide the reference records whose production data belongs to the same type of production data set among all the reference records of the target call record into a set to obtain a plurality of reference record sets;

[0074] The similar record generation probability calculation unit is used to obtain the number U of data call records whose correlation degree between the production data corresponding to the target call record is not 0 in the production data storage service system. g , if 1+U g <U r , then the corresponding reference record set r is set as the valid set, and the ratio of the number of valid sets to the number of reference record sets is calculated as the probability η that the target call record will subsequently generate the same type of data call record; where U r is the number of reference records in the reference record set r, and the same type of data call record represents the data call record in the same type of data record set;

[0075] The expected interval acquisition unit is used to convert M g Perform similarity comparison with the feature information set of all data call records in all valid sets, select the data call record with the greatest similarity as the most similar data call record, and extract the data call record whose corresponding production data's most recent call time is after the most similar data call record from the valid set where the most similar data call record is located, and use the interval between the most recent call times of the production data of these two data call records as the expected time interval Δt for the subsequent generation of similar data call records by the target call record;

[0076] The expected interface feedback calculation unit is used to calculate the expected interface feedback value of the target call record after the expected time interval Δt according to the expected time interval Δt and the probability η according to the following formula:

[0077]

[0078] In the formula, (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt.

[0079] Furthermore, the abnormality identification module specifically includes:

[0080] The actual interface feedback calculation unit is used to obtain the calling frequency P of the production data corresponding to the target calling record after the expected time interval Δt. nowand the amount of data called Q now , and calculate the actual interface feedback value (F) of the target call record after the expected time interval Δt according to the interface feedback model g ) ac ;

[0081] A determination unit, used to determine whether a new data call record of the same type is generated after the target call record in the production data storage service system;

[0082] The exception identification unit is used to generate a new data call record of the same type, and Then an abnormal reminder is given for the calling behavior recorded in the target calling record, wherein (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt. Indicates the real-time interface feedback value of the target call record.

[0083] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above method.

[0084] A computer-readable storage medium having a computer program / instruction stored thereon, wherein the computer program / instruction implements the above method when executed by a processor.

[0085] Compared with the prior art, the present invention has the following beneficial effects: the present invention constructs an interface feedback model to analyze the data content of the production data of the pumped-storage power station and the downstream interface call situation, integrates the actual data content and data consumption situation, avoids the adverse effects caused by obtaining less information through data content, and improves the accuracy of the system in identifying abnormal calls to the production data of the pumped-storage power station. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 It is a flow chart of a method for identifying abnormalities in calling production data of a pumped storage power station provided by an embodiment of the present invention;

[0087] Figure 2 is a structural diagram of a system for identifying abnormalities in calling production data of a pumped-storage power station provided by an embodiment of the present invention;

[0088] Figure 3 It is a structural diagram of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0089] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0090] Embodiment 1

[0091] The embodiment of the present invention provides a method for identifying abnormal call of production data of a pumped storage power station, such as Figure 1 As shown, the following steps are included.

[0092] Step S100: obtaining data call records of the application calling the production data of the pumped storage power station through the interface, and obtaining high-frequency call fields in the corresponding production data based on the data call records, and using the high-frequency call fields together with the production data itself as a feature set of the corresponding production data.

[0093] Each data call record is used to record all called data areas and call times in the production data of a production system of a pumped-storage power station.

[0094] Step S100 specifically includes:

[0095] Step S101: Acquire all called data with call records in the production data stored in the production data storage service system, count the occurrence frequency of each field of all called data, and store each field and the corresponding occurrence frequency in the data content database; wherein the production data and data call records of the pumped storage power station are stored in the production data storage service system;

[0096] Step S102: extract all data call records from the production data storage service system, divide the corresponding production data of each data call record into several information blocks, compare the fields in each information block with the fields stored in the data content database, if the fields of the information block and the fields stored in the data content database have the same fields, select the field with the highest frequency of occurrence from the same fields as the characteristic field of the corresponding information block, and use the information block itself as the characteristic information of the corresponding information block; if there are no identical fields, directly use the information block itself as the characteristic information of the corresponding information block;

[0097] Step S103: The set of characteristic fields and characteristic information of all information blocks of the production data corresponding to the data call record is taken as its characteristic set.

[0098] Step S200: calculating the correlation between each production data according to the similarity of the feature set of the production data, and classifying the production data with a correlation greater than a threshold into a similar production data set.

[0099] The step S200 specifically includes:

[0100] Step S201: For any data call record i and any data call record j, respectively calculate the number of identical feature fields H in the feature field set of production data corresponding to the data call record i and the feature field set of production data corresponding to the data call record j ij , and the number of identical feature fields H ij Divide by the maximum number of fields in the two feature field sets to get the proportion of the same feature fields in the production data corresponding to data call record i and data call record j X ij ;

[0101] The basis for determining the same field is as follows: the similarity between the two fields is calculated. If the similarity exceeds the set similarity threshold, the two compared feature fields are set as the same field. For example, the first set of feature field sets contains the word "power generation", and the second set of feature field sets contains the word "power generation amount". The two fields are compared for similarity. Since the similarity exceeds the set similarity threshold, "power generation amount" and "power generation amount" can be set as the same field.

[0102] Step S202: for any data call record i and any data call record j, calculate the similarity between each feature information of the production data corresponding to the data call record i and each feature information of the production data corresponding to the data call record j, and if the similarity is less than the similarity threshold β, assign the corresponding similarity to 0; wherein the similarity between the feature information is calculated using a text similarity calculation method, such as the longest common subsequence (LCS) method, the edit distance method, the cosine similarity method, etc.;

[0103] Step S203: For any data call record i and any data call record j, the similarity S of the feature information set of the production data corresponding to the data call record i and the data call record j is calculated according to the following formula: ij ;

[0104]

[0105] In the formula, (S ij ) ab N represents the similarity between the ath feature information of the production data corresponding to the data call record i and the bth feature information of the production data corresponding to the data call record j, i 、N j They represent the number of characteristic information of production data corresponding to data call records i and j, respectively, and Max() is the maximum value function;

[0106] Step S204: If X ij <γ, then the correlation G between the data call records i and j corresponding to the production data ij Assign a value of 0; if X ij≥γ, then G ij =X ij ×(1+S ij ), where γ is the preset proportion threshold;

[0107] Step S205: If G ij ≠0, the production data corresponding to the data call records i and j are classified into a category of data, and the data call record i and the data call record j are generated into a set of data call records of the same category.

[0108] Step S300: Calculate the calling frequency and calling data volume of production data at each moment according to the called data area and calling time recorded in the data calling record.

[0109] The data call record records the data area and call time of each call. By summarizing and calculating this data, the call frequency and call data volume of the production data at each moment can be obtained.

[0110] Step S400: construct and train an interface feedback model.

[0111] Among them, the interface feedback model calculates the interface feedback value of the corresponding data call record according to the calling frequency, calling data volume and correlation between production data at each moment, and the interface feedback value is positively correlated with the calling frequency, calling data volume and correlation between production data.

[0112] Specifically, step S400 includes the following steps:

[0113] Step S401: constructing an interface feedback model:

[0114]

[0115] In the formula, (F i ) t represents the interface feedback value of the current data call record i at time t, y is the coefficient to be trained, (P i ) t , (Q i ) t They represent the calling frequency and calling data volume of the production data corresponding to the data calling record i at time t, G ki Indicates the correlation between the production data corresponding to data call record k and data call record i, where the index number of the data call record is assigned in sequence starting from 1 according to the generation time, the generation time of data call record k is before the generation time of data call record i, and t represents the tth unit time after the production time;

[0116] Step S402: Select several data call records from the production data storage service system as samples, and for each sample, obtain other data call records whose generation time is before the sample, and obtain the correlation between the production data corresponding to these data call records and the production data corresponding to the current data call record;

[0117] Step S403: Calculate the difference between the total amount of data at any two adjacent moments of each sample, where the total amount of data is the product of the call frequency and the call data amount;

[0118] Step S404: With the goal of making the difference between the total amount of data at any two adjacent moments of each sample consistent with the change of the corresponding interface feedback value, the interface feedback model is trained to obtain the coefficient y to be trained, and the training of the interface feedback model is completed. Wherein, in the case of an uncertain constant coefficient y, the interface feedback value obtained by the interface feedback model is a value with y, and in the same data call record, because the correlation with the other data call records is the same, the interface feedback value is only affected by the call frequency and the call amount; when the feedback data of the interface does not increase much over time, it means that the frequency of use of the production data of the pumped storage power station has decreased, so the interface feedback value should also decrease, so the constant coefficient y is needed to match the change between the feedback data and the interface feedback value. During training, y only needs to be able to keep the change of the difference in the total amount of data consistent with the change law of the corresponding interface feedback value.

[0119] Step S500: Set the data call record with the most recent generation time to the current moment as the target call record, and calculate the real-time interface feedback value of the target call record according to the interface feedback model, obtain the data call record whose production data corresponding to the target call record belongs to the same production data set, and calculate the probability of the target call record subsequently generating the same type of data call record based on these data call records, as well as the expected interface feedback value of the target call record after a period of time.

[0120] Step S500 specifically includes:

[0121] Step S501: Take the data call record with the closest generation time to the current time as the target call record, and obtain the call frequency P of the production data corresponding to the target call record g And data call volume Q g , and the real-time interface feedback value (F g ) tg :

[0122] Step S502: Obtain the feature set T of the production data corresponding to the target call record g =(W g,M g ), where W g The characteristic field set of the target call record corresponding to the production data, M g The characteristic information set of the corresponding production data is recorded for the target call, and W g Compare it with the characteristic field set of production data corresponding to all other data call records, and compare it with W g The same target call record is set as a reference record of the target call record;

[0123] Step S503: among all reference records of the target call record, the reference records whose production data belong to the same type of production data set are divided into one set to obtain a plurality of reference record sets;

[0124] Step S504: Obtain the number of data call records whose correlation degree with the production data corresponding to the target call record is not 0 in the production data storage service system. g , if 1+U g <U r , then the corresponding reference record set r is set as the valid set, and the ratio of the number of valid sets to the number of reference record sets is calculated as the probability η that the target call record will subsequently generate the same type of data call record; where U r is the number of reference records in the reference record set r, and the same type of data call record represents the data call record in the same type of data record set;

[0125] Step S505: M g Perform similarity comparison with the feature information set of all data call records in all valid sets, select the data call record with the greatest similarity as the most similar data call record, and extract the data call record whose corresponding production data's most recent call time is after the most similar data call record from the valid set where the most similar data call record is located, and use the interval between the most recent call times of the production data of these two data call records as the expected time interval Δt for the subsequent generation of similar data call records by the target call record;

[0126] Step S506: According to the expected time interval Δt and the probability η, the expected interface feedback value of the target call record after the expected time interval Δt has passed is calculated according to the following formula:

[0127]

[0128] In the formula, (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt.

[0129] Step S600: Calculate the actual interface feedback value of the target call record after the period of time according to the interface feedback model, and determine whether to issue an abnormal reminder for the call behavior corresponding to the target call record by comparing the difference between the actual interface feedback value of the target call record and the expected interface feedback value.

[0130] Step S600 specifically includes:

[0131] Step S601: Obtain the calling frequency P of the production data corresponding to the target calling record after the expected time interval Δt now and the amount of data called Q now , and calculate the actual interface feedback value (F) of the target call record after the expected time interval Δt according to the interface feedback model g ) ac ;

[0132] Step S602: determining whether a new data call record of the same type is generated after the target call record in the production data storage service system;

[0133] Step S603: If a new data call record of the same type is generated, and Then an abnormal reminder is given for the calling behavior recorded in the target calling record, wherein (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt. Indicates the real-time interface feedback value of the target call record.

[0134] Embodiment 2

[0135] The embodiment of the present invention provides a system for identifying abnormal calls of production data of a pumped storage power station, such as Figure 2 As shown, including:

[0136] A feature extraction module is used to obtain data call records of the application calling the production data of the pumped storage power station through the interface, and based on the data call records, obtain the high-frequency call fields in the corresponding production data, and use them together with the production data itself as the feature set of the corresponding production data; wherein each data call record is used to record all called data areas and call times in the production data of a production system of the pumped storage power station;

[0137] A classification module is used to calculate the correlation between each production data according to the similarity of the feature set of the production data, and classify the production data with a correlation greater than a threshold into a similar production data set;

[0138] The calling feature calculation module is used to calculate the calling frequency and calling data volume of the production data at each moment according to the called data area and calling time recorded in the data calling record;

[0139] A model training module is used to construct and train an interface feedback model, wherein the interface feedback model calculates the interface feedback value of the corresponding data call record according to the call frequency, call data volume and correlation between production data at each moment, wherein the interface feedback value is positively correlated with the call frequency, call data volume and correlation between production data;

[0140] The expected interface feedback calculation module is used to set the data call record with the closest generation time to the current time as the target call record, and calculate the real-time interface feedback value of the target call record according to the interface feedback model, obtain the data call record whose production data corresponding to the target call record belongs to the same production data set, and calculate the probability of the target call record generating the same type of data call record later according to these data call records, as well as the expected interface feedback value of the target call record after a period of time;

[0141] The abnormality identification module is used to calculate the actual interface feedback value of the target call record after the period of time according to the interface feedback model, and determine whether to issue an abnormal reminder for the call behavior corresponding to the target call record by comparing the difference between the actual interface feedback value of the target call record and the expected interface feedback value.

[0142] Wherein, the feature extraction module specifically includes:

[0143] A high-frequency call field statistics unit is used to obtain all called data with call records in the production data stored in the production data storage service system, count the occurrence frequency of each field of all called data, and store each field and the corresponding occurrence frequency in the data content database; wherein the production data and data call records of the pumped storage power station are stored in the production data storage service system;

[0144] The feature extraction unit is used to extract all data call records from the production data storage service system. For each data call record, the corresponding production data is divided into several information blocks, and each field in each information block is compared with the fields stored in the data content database. If the fields of the information block and the fields stored in the data content database are identical, the field with the highest frequency of occurrence is selected from the identical fields as the feature field of the corresponding information block, and the information block itself is used as the feature information of the corresponding information block. If there are no identical fields, the information block itself is directly used as the feature information of the corresponding information block.

[0145] The feature forming unit is used to take the set of feature fields and feature information of all information blocks of the production data corresponding to the data call record as its feature set.

[0146] Wherein, the classification module specifically includes:

[0147] The same field calculation unit is used to calculate the number of the same feature fields in the feature field set of the production data corresponding to the data call record i and the feature field set of the production data corresponding to the data call record j for any data call record i and any data call record j, respectively. ij , and the number of identical feature fields H ij Divide by the maximum number of fields in the two feature field sets to get the proportion of the same feature fields in the production data corresponding to data call record i and data call record j X ij ;

[0148] The information block similarity calculation unit is used to calculate the similarity between each feature information of the production data corresponding to the data call record i and each feature information of the production data corresponding to the data call record j for any data call record i and any data call record j, and if the similarity is less than a similarity threshold β, the corresponding similarity is assigned to 0;

[0149] The feature information similarity calculation unit is used to calculate the similarity S of the feature information set of the production data corresponding to the data call record i and the data call record j according to the following formula for any data call record i and any data call record j ij ;

[0150]

[0151] In the formula, (S ij ) ab N represents the similarity between the ath feature information of the production data corresponding to the data call record i and the bth feature information of the production data corresponding to the data call record j, i 、N j They represent the number of characteristic information of production data corresponding to data call records i and j, respectively, and Max() is the maximum value function;

[0152] The correlation calculation unit is used if X ij <γ, then the correlation G between the data call records i and j corresponding to the production data ij Assign a value of 0; if X ij ≥γ, then G ij =X ij ×(1+S ij ), where γ is the preset proportion threshold;

[0153] Classification unit, used if Gij ≠0, the production data corresponding to the data call records i and j are classified into a category of data, and the data call record i and the data call record j are generated into a set of data call records of the same category.

[0154] The model training module specifically includes:

[0155] The model building unit, used to build the interface feedback model is:

[0156]

[0157] In the formula, (F i ) t represents the interface feedback value of the current data call record i at time t, y is the coefficient to be trained, (P i ) t , (Q i ) t They represent the calling frequency and calling data volume of the production data corresponding to the data calling record i at time t, G ki Indicates the correlation between the production data corresponding to data call record k and data call record i, where the index number of the data call record is assigned in sequence starting from 1 according to the generation time, the generation time of data call record k is before the generation time of data call record i, and t represents the tth unit time after the production time;

[0158] A sample selection unit is used to select a number of data call records from the production data storage service system as samples, and for each sample, obtain other data call records whose generation time is before the sample, and obtain the correlation between the production data corresponding to these data call records and the production data corresponding to the current data call record;

[0159] A data total amount calculation unit, used to calculate the difference between the total amount of data of each sample at any two adjacent moments, wherein the total amount of data is the product of the call frequency and the call data amount;

[0160] The model training unit is used to train the interface feedback model with the goal of making the change in the difference in the total amount of data between any two adjacent moments of each sample consistent with the change in the corresponding interface feedback value, obtain the coefficient y to be trained, and complete the training of the interface feedback model.

[0161] The expected interface feedback calculation module specifically includes:

[0162] The real-time interface sensing calculation unit is used to take the data call record with the closest generation time to the current time as the target call record, and obtain the call frequency P of the target call record corresponding to the production data g And data call volume Q g, and calculate the real-time interface feedback value of the target call record according to the following formula based on the interface feedback model

[0163] Reference record acquisition unit, used to obtain the feature set T of the production data corresponding to the target call record g =(W g ,M g ), where W g The characteristic field set of the target call record corresponding to the production data, M g The characteristic information set of the corresponding production data is recorded for the target call, and W g Compare it with the characteristic field set of production data corresponding to all other data call records, and compare it with W g The same target call record is set as a reference record of the target call record;

[0164] A reference record set acquisition unit is used to divide the reference records whose production data belongs to the same type of production data set among all the reference records of the target call record into a set to obtain a plurality of reference record sets;

[0165] The similar record generation probability calculation unit is used to obtain the number U of data call records whose correlation degree between the production data corresponding to the target call record is not 0 in the production data storage service system. g , if 1+U g <U r , then the corresponding reference record set r is set as the valid set, and the ratio of the number of valid sets to the number of reference record sets is calculated as the probability η that the target call record will subsequently generate the same type of data call record; where U r is the number of reference records in the reference record set r, and the same type of data call record represents the data call record in the same type of data record set;

[0166] The expected interval acquisition unit is used to convert M g Perform similarity comparison with the feature information set of all data call records in all valid sets, select the data call record with the greatest similarity as the most similar data call record, and extract the data call record whose corresponding production data's most recent call time is after the most similar data call record from the valid set where the most similar data call record is located, and use the interval between the most recent call times of the production data of these two data call records as the expected time interval Δt for the subsequent generation of similar data call records by the target call record;

[0167] The expected interface feedback calculation unit is used to calculate the expected interface feedback value of the target call record after the expected time interval Δt according to the expected time interval Δt and the probability η according to the following formula:

[0168]

[0169] In the formula, (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt.

[0170] The abnormality identification module specifically includes:

[0171] The actual interface feedback calculation unit is used to obtain the calling frequency P of the production data corresponding to the target calling record after the expected time interval Δt. now and the amount of data called Q now , and calculate the actual interface feedback value (F) of the target call record after the expected time interval Δt according to the interface feedback model g ) ac ;

[0172] A determination unit, used to determine whether a new data call record of the same type is generated after the target call record in the production data storage service system;

[0173] The exception identification unit is used to generate a new data call record of the same type, and Then an abnormal reminder is given for the calling behavior recorded in the target calling record, wherein (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt. Indicates the real-time interface feedback value of the target call record.

[0174] The system provided in the embodiment of the present invention can be used to execute the method provided in the first embodiment of the present invention, and has the corresponding functions and beneficial effects of the execution method.

[0175] It is worth noting that in the embodiment of the above system, the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0176] The embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, or of course, it can be implemented only by hardware, as long as the function or effect can be achieved.

[0177] Embodiment 3

[0178] Figure 3 Schematic diagram of the structure of a computer device provided by an embodiment of the present invention. The embodiment of the present invention provides services for implementing the method of the above-mentioned embodiment 1 of the present invention. Figure 3 As shown, the device may include: a memory 301 storing a computer executable program; a processor 302 coupled to the memory 301; the processor 302 calls the computer executable program stored in the memory 301 to execute the steps in the method described in the first embodiment.

[0179] The memory 301 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) and / or cache memory. The device may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the memory 301 may be used to read and write non-removable, non-volatile magnetic media (commonly referred to as "hard drives"). A program / utility having a set (at least one) of program modules may be stored in, for example, the memory 301, such program modules including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include the implementation of a network environment. The computer executable program of the program module typically performs the functions and / or methods in the embodiments described in the present invention.

[0180] The processor 302 executes various functional applications and data processing by running the program stored in the memory 301, such as implementing the method provided in the first embodiment of the present invention.

[0181] The code of the computer executable program can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" language or similar programming languages.

[0182] Embodiment 4

[0183] An embodiment of the present invention provides a storage medium including a computer executable program. When the computer executable program is executed by a computer processor, it is used to execute the method of the first embodiment.

[0184] The storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. Computer-readable media can be computer-readable signal media or computer-readable storage media. Computer-readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: electrical connections with one or more wires, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this document, computer-readable storage media can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0185] Of course, the storage medium containing a computer executable program provided by an embodiment of the present invention, the computer executable program of which is not limited to the above method operations, can also execute related operations in the method provided by any embodiment of the present invention.

[0186] It should be understood that the above embodiments and descriptions only describe the principles, main features and advantages of the present invention. Without departing from the spirit and scope of the present invention, the present invention may be subject to various changes and improvements, and these changes and improvements all fall within the scope of protection of the present invention.

Claims

1. A method for identifying abnormal call of production data of a pumped storage power station, characterized in that: include: Step S100: obtaining data call records of the application calling the production data of the pumped-storage power station through the interface, and obtaining high-frequency call fields in the corresponding production data based on the data call records, and using the high-frequency call fields together with the production data itself as a feature set of the corresponding production data; wherein each data call record is used to record all called data areas and call times in the production data of a production system of the pumped-storage power station; Step S200: calculating the correlation between each production data according to the similarity of the feature set of the production data, and classifying the production data with a correlation greater than a threshold into a similar production data set; Step S300: Calculate the calling frequency and calling data volume of production data at each moment according to the called data area and calling time recorded in the data calling record; Step S400: constructing and training an interface feedback model, wherein the interface feedback model calculates the interface feedback value of the corresponding data call record according to the call frequency, call data volume and correlation between production data at each moment, wherein the interface feedback value is positively correlated with the call frequency, call data volume and correlation between production data; Step S500: Set the data call record whose generation time is closest to the current moment as the target call record, and calculate the real-time interface feedback value of the target call record according to the interface feedback model, obtain the data call record whose production data corresponding to the target call record belongs to the same production data set, and calculate the probability that the target call record will subsequently generate the same type of data call record according to these data call records, as well as the expected interface feedback value of the target call record after a period of time; Step S600: Calculate the actual interface feedback value of the target call record after the period of time according to the interface feedback model, and determine whether to issue an abnormal reminder for the call behavior corresponding to the target call record by comparing the difference between the actual interface feedback value of the target call record and the expected interface feedback value.

2. The method for identifying abnormal call of production data of a pumped storage power station according to claim 1, characterized in that: The step S100 specifically includes: Step S101: Acquire all called data with call records in the production data stored in the production data storage service system, count the occurrence frequency of each field of all called data, and store each field and the corresponding occurrence frequency in the data content database; wherein the production data and data call records of the pumped storage power station are stored in the production data storage service system; Step S102: extract all data call records from the production data storage service system, divide the corresponding production data of each data call record into several information blocks, compare the fields in each information block with the fields stored in the data content database, if the fields of the information block and the fields stored in the data content database have the same fields, select the field with the highest frequency of occurrence from the same fields as the characteristic field of the corresponding information block, and use the information block itself as the characteristic information of the corresponding information block; if there are no identical fields, directly use the information block itself as the characteristic information of the corresponding information block; Step S103: The set of characteristic fields and characteristic information of all information blocks of the production data corresponding to the data call record is taken as its characteristic set.

3. The method for identifying abnormal call of production data of a pumped storage power station according to claim 2, characterized in that: The step S200 specifically includes: Step S201: For any data call record i and any data call record j, respectively calculate the number of identical feature fields H in the feature field set of production data corresponding to the data call record i and the feature field set of production data corresponding to the data call record j ij , and the number of identical feature fields H ij Divide by the maximum number of fields in the two feature field sets to get the proportion of the same feature fields in the production data corresponding to data call record i and data call record j X ij ; Step S202: for any data call record i and any data call record j, calculate the similarity between each feature information of the production data corresponding to the data call record i and each feature information of the production data corresponding to the data call record j, and if the similarity is less than the similarity threshold β, assign the corresponding similarity to 0; Step S203: For any data call record i and any data call record j, the similarity S of the feature information set of the production data corresponding to the data call record i and the data call record j is calculated according to the following formula: ij ;; In the formula, (S ij ) ab N represents the similarity between the ath feature information of the production data corresponding to the data call record i and the bth feature information of the production data corresponding to the data call record j, i 、N j They represent the number of characteristic information of production data corresponding to data call records i and j, respectively, and Max() is the maximum value function; Step S204: If X ij <γ, then the correlation G between the data call records i and j corresponding to the production data ij Assign a value of 0; if X ij ≥γ, then G ij =X ij ×(1+S ij ), where γ is the preset proportion threshold; Step S205: If G ij ≠0, the production data corresponding to the data call records i and j are classified into a category of data, and the data call record i and the data call record j are generated into a set of data call records of the same category.

4. The method for identifying abnormal call of production data of a pumped storage power station according to claim 1, characterized in that: The step S400 specifically includes: Step S401: constructing an interface feedback model: In the formula, (F i ) t represents the interface feedback value of the current data call record i at time t, y is the coefficient to be trained, (P i ) t , (Q i ) t They represent the calling frequency and calling data volume of the production data corresponding to the data calling record i at time t, G ki Indicates the correlation between the production data corresponding to data call record k and data call record i, where the index number of the data call record is assigned in sequence starting from 1 according to the generation time, the generation time of data call record k is before the generation time of data call record i, and t represents the tth unit time after the production time; Step S402: Select several data call records from the production data storage service system as samples, and for each sample, obtain other data call records whose generation time is before the sample, and obtain the correlation between the production data corresponding to these data call records and the production data corresponding to the current data call record; Step S403: Calculate the difference between the total amount of data at any two adjacent moments of each sample, where the total amount of data is the product of the call frequency and the call data amount; Step S404: With the goal of making the difference between the total amount of data at any two adjacent moments of each sample consistent with the change in the corresponding interface feedback value, the interface feedback model is trained to obtain the coefficient y to be trained, thereby completing the training of the interface feedback model.

5. The method for identifying abnormal call of production data of a pumped storage power station according to claim 1, characterized in that: The step S500 specifically includes: Step S501: Take the data call record with the closest generation time to the current time as the target call record, and obtain the call frequency P of the production data corresponding to the target call record g And data call volume Q g , and the real-time interface feedback value (F g ) tg : Step S502: Obtain the feature set T of the production data corresponding to the target call record g =(W g ,M g ), where W g The characteristic field set of the target call record corresponding to the production data, M g The characteristic information set of the corresponding production data is recorded for the target call, and W g Compare it with the characteristic field set of production data corresponding to all other data call records, and compare it with W g The same target call record is set as a reference record of the target call record; Step S503: among all reference records of the target call record, the reference records whose production data belong to the same type of production data set are divided into one set to obtain a plurality of reference record sets; Step S504: Obtain the number of data call records whose correlation degree with the production data corresponding to the target call record is not 0 in the production data storage service system. g , if 1+U g <U r , then the corresponding reference record set r is set as the valid set, and the ratio of the number of valid sets to the number of reference record sets is calculated as the probability η that the target call record will subsequently generate the same type of data call record; where U r is the number of reference records in the reference record set r, and the same type of data call record represents the data call record in the same type of data record set; Step S505: M g Perform similarity comparison with the feature information set of all data call records in all valid sets, select the data call record with the greatest similarity as the most similar data call record, and extract the data call record whose corresponding production data's most recent call time is after the most similar data call record from the valid set where the most similar data call record is located, and use the interval between the most recent call times of the production data of these two data call records as the expected time interval Δt for the subsequent generation of similar data call records by the target call record; Step S506: According to the expected time interval Δt and the probability η, the expected interface feedback value of the target call record after the expected time interval Δt has passed is calculated according to the following formula: In the formula, (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt.

6. The method for identifying abnormal call of production data of a pumped storage power station according to claim 1, characterized in that: The step S600 specifically includes: Step S601: Obtain the calling frequency P of the production data corresponding to the target calling record after the expected time interval Δt now and the amount of data called Q now , and calculate the actual interface feedback value (F) of the target call record after the expected time interval Δt according to the interface feedback model g ) ac ; Step S602: determining whether a new data call record of the same type is generated after the target call record in the production data storage service system; Step S603: If a new data call record of the same type is generated, and Then an abnormal reminder is given for the calling behavior recorded in the target calling record, wherein (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt. Indicates the real-time interface feedback value of the target call record.

7. A system for identifying abnormal calls of production data of a pumped storage power station, characterized in that: include: A feature extraction module is used to obtain data call records of the application calling the production data of the pumped storage power station through the interface, and based on the data call records, obtain the high-frequency call fields in the corresponding production data, and use them together with the production data itself as the feature set of the corresponding production data; wherein each data call record is used to record all called data areas and call times in the production data of a production system of the pumped storage power station; A classification module is used to calculate the correlation between each production data according to the similarity of the feature set of the production data, and classify the production data with a correlation greater than a threshold into a similar production data set; The calling feature calculation module is used to calculate the calling frequency and calling data volume of the production data at each moment according to the called data area and calling time recorded in the data calling record; A model training module is used to construct and train an interface feedback model, wherein the interface feedback model calculates the interface feedback value of the corresponding data call record according to the call frequency, call data volume and correlation between production data at each moment, wherein the interface feedback value is positively correlated with the call frequency, call data volume and correlation between production data; The expected interface feedback calculation module is used to set the data call record with the closest generation time to the current time as the target call record, and calculate the real-time interface feedback value of the target call record according to the interface feedback model, obtain the data call record whose production data corresponding to the target call record belongs to the same production data set, and calculate the probability of the target call record generating the same type of data call record later according to these data call records, as well as the expected interface feedback value of the target call record after a period of time; The abnormality identification module is used to calculate the actual interface feedback value of the target call record after the period of time according to the interface feedback model, and determine whether to issue an abnormal reminder for the call behavior corresponding to the target call record by comparing the difference between the actual interface feedback value of the target call record and the expected interface feedback value.

8. The system for identifying abnormality in calling production data of a pumped storage power station according to claim 7, characterized in that: The feature extraction module specifically includes: A high-frequency call field statistics unit is used to obtain all called data with call records in the production data stored in the production data storage service system, count the occurrence frequency of each field of all called data, and store each field and the corresponding occurrence frequency in the data content database; wherein the production data and data call records of the pumped storage power station are stored in the production data storage service system; The feature extraction unit is used to extract all data call records from the production data storage service system. For each data call record, the corresponding production data is divided into several information blocks, and each field in each information block is compared with the fields stored in the data content database. If the fields of the information block and the fields stored in the data content database are identical, the field with the highest frequency of occurrence is selected from the identical fields as the feature field of the corresponding information block, and the information block itself is used as the feature information of the corresponding information block. If there are no identical fields, the information block itself is directly used as the feature information of the corresponding information block. The feature forming unit is used to take the set of feature fields and feature information of all information blocks of the production data corresponding to the data call record as its feature set.

9. The system for identifying abnormalities in calling production data of a pumped storage power station according to claim 8, characterized in that: The classification module specifically includes: The same field calculation unit is used to calculate the number of the same feature fields H in the feature field set of the production data corresponding to the data call record i and the feature field set of the production data corresponding to the data call record j for any data call record i and any data call record j respectively. ij , and the number of identical feature fields H ij Divide by the maximum number of fields in the two feature field sets to get the proportion of the same feature fields in the production data corresponding to data call record i and data call record j X ij ; The information block similarity calculation unit is used to calculate the similarity between each feature information of the production data corresponding to the data call record i and each feature information of the production data corresponding to the data call record j for any data call record i and any data call record j, and if the similarity is less than a similarity threshold β, the corresponding similarity is assigned to 0; The feature information similarity calculation unit is used to calculate the similarity S of the feature information set of the production data corresponding to the data call record i and the data call record j according to the following formula for any data call record i and any data call record j ij ; In the formula, (S ij ) ab N represents the similarity between the ath feature information of the production data corresponding to the data call record i and the bth feature information of the production data corresponding to the data call record j, i 、N j They represent the number of characteristic information of production data corresponding to data call records i and j, respectively, and Max() is the maximum value function; The correlation calculation unit is used if X ij <γ, then the correlation G between the data call records i and j corresponding to the production data ij Assign a value of 0; if X ij ≥γ, then G ij =X ij ×(1+S ij ), where γ is the preset proportion threshold; Classification unit, used if G ij ≠0, the production data corresponding to the data call records i and j are classified into a category of data, and the data call record i and the data call record j are generated into a set of data call records of the same category.

10. The system for identifying abnormal call of production data of a pumped storage power station according to claim 7, characterized in that: The model training module specifically includes: The model building unit, used to build the interface feedback model is: In the formula, (F i ) t represents the interface feedback value of the current data call record i at time t, y is the coefficient to be trained, (P i ) t , (Q i ) t They represent the calling frequency and calling data volume of the production data corresponding to the data calling record i at time t, G ki Indicates the correlation between the production data corresponding to data call record k and data call record i, where the index number of the data call record is assigned in sequence starting from 1 according to the generation time, the generation time of data call record k is before the generation time of data call record i, and t represents the tth unit time after the production time; A sample selection unit is used to select a number of data call records from the production data storage service system as samples, and for each sample, obtain other data call records whose generation time is before the sample, and obtain the correlation between the production data corresponding to these data call records and the production data corresponding to the current data call record; A data total amount calculation unit, used to calculate the difference between the total amount of data of each sample at any two adjacent moments, wherein the total amount of data is the product of the call frequency and the call data amount; The model training unit is used to train the interface feedback model with the goal of making the change in the difference in the total amount of data between any two adjacent moments of each sample consistent with the change in the corresponding interface feedback value, obtain the coefficient y to be trained, and complete the training of the interface feedback model.

11. The system for identifying abnormality in calling production data of a pumped storage power station according to claim 7, characterized in that: The expected interface feedback calculation module specifically includes: The real-time interface sensing calculation unit is used to take the data call record with the closest generation time to the current time as the target call record, and obtain the call frequency P of the target call record corresponding to the production data g And data call volume Q g , and calculate the real-time interface feedback value of the target call record according to the following formula based on the interface feedback model Reference record acquisition unit, used to obtain the feature set T of the production data corresponding to the target call record g =(W g ,M g ), where W g The characteristic field set of the target call record corresponding to the production data, M g The characteristic information set of the corresponding production data is recorded for the target call, and W g Compare it with the characteristic field set of production data corresponding to all other data call records, and compare it with W g The same target call record is set as a reference record of the target call record; A reference record set acquisition unit is used to divide the reference records whose production data belongs to the same production data set among all the reference records of the target call record into a set to obtain a plurality of reference record sets; The similar record generation probability calculation unit is used to obtain the number U of data call records whose correlation degree between the production data corresponding to the target call record is not 0 in the production data storage service system. g , if 1+U g <U r , then the corresponding reference record set r is set as the valid set, and the ratio of the number of valid sets to the number of reference record sets is calculated as the probability η that the target call record will subsequently generate the same type of data call record; where U r is the number of reference records in the reference record set r, and the same type of data call record represents the data call record in the same type of data record set; The expected interval acquisition unit is used to convert M g Perform similarity comparison with the feature information set of all data call records in all valid sets, select the data call record with the greatest similarity as the most similar data call record, and extract the data call record whose corresponding production data's most recent call time is after the most similar data call record from the valid set where the most similar data call record is located, and use the interval between the most recent call times of the production data of these two data call records as the expected time interval Δt for the subsequent generation of similar data call records by the target call record; The expected interface feedback calculation unit is used to calculate the expected interface feedback value of the target call record after the expected time interval Δt according to the expected time interval Δt and the probability η according to the following formula: In the formula, (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt.

12. The system for identifying abnormality in calling production data of a pumped storage power station according to claim 7, characterized in that: The abnormality identification module specifically includes: The actual interface feedback calculation unit is used to obtain the calling frequency P of the production data corresponding to the target calling record after the expected time interval Δt. now and the amount of data called Q now , and calculate the actual interface feedback value (F) of the target call record after the expected time interval Δt according to the interface feedback model g ) ac ; A determination unit, used to determine whether a new data call record of the same type is generated after the target call record in the production data storage service system; The exception identification unit is used to generate a new data call record of the same type, and Then an abnormal reminder is given for the calling behavior recorded in the target calling record, wherein (F g ) ex It represents the expected interface feedback value of the target call record after the expected time interval Δt. Indicates the real-time interface feedback value of the target call record.

13. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the computer program to implement the method according to any one of claims 1 to 6.

14. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: The computer program / instructions, when executed by a processor, implement the method of any one of claims 1-6.

Images